Operators of cellular telecommunications networks are ever trying to improve the efficiency of the networks and optimise network performance as poor customer experience and service loss can result in customer dissatisfaction. In conventional cellular telecommunications networks, when the user is requesting a data service, such as downloading an application or streaming audio or video data, the quality of service provided to that user is directly related to network conditions for example radio coverage and delay. If the user moves into a white spot while requesting data service, for example while travelling in a car or on a train, the data service may be interrupted (i.e. will stop receiving the data service until coverage is recovered) or may be lost completely as the data link may be damaged. A white spot is a term used in the art to describe a geographical region lacking radio coverage. Such a loss of data service results in customer dissatisfaction and often customer complaints. There is therefore a need to reduce the effects of poor service coverage on the end user in order to improve the customer experience.
It has previously been described, for example in European Patent Publication No. 2,403,290, that a network operator may be able to gauge future radio conditions experienced by a user by prediction. In order to make this prediction, the network may build and maintain a record of radio quality across a coverage area of the network and may use this record to predict the radio conditions available to the mobile terminals within that coverage area. The network will typically calculate the present location and/or velocity of the user in order to predict the future location of that user. By consulting the record of radio quality across the coverage area the network is able to estimate the radio conditions available to that mobile terminal at a future location.
If it is determined that future radio conditions are poor then the network may increase the data throughput provided to that user at its present location. The user can then store this received information in a buffer so that the data is temporarily stored for future use when the radio conditions are poor. Alternatively, if it is determined that future radio conditions are good, then the network may decrease the data throughput provided to that user at its present location since the user will not need to use stored data at the future location in order to receive a constant service.
Additionally, it is known from for example European Patent Publication No. 2,555,569, that network parameters (such as resource allocation and resource prioritisation, radio parameters and Quality of Service, QoS, parameters) and service parameters (such as resolution and time of delivery) can be adjusted for a specific data service based on the future radio conditions determined using the current location and velocity of the user. Predicting the future location of a user based on current location and velocity in this way is highly inaccurate as a simple left turn will dramatically alter the predicted destination. The future location could be recalculated after this turn but this would impose large overheads and require constant iterative recalculation. There would also be only a period of time in the future that the location could be predicted for. There can be no accurate long term prediction of coverage.
In U.S. Pat. No. 6,125,278 it has been described that location data can be sent from the user to the network and combined with current location and velocity data to predict the future location of the user. Whilst this may increase the accuracy of location prediction over the use of current location and velocity data alone, the inaccuracy of this method is still undesirable and the use of device information in this way introduces undesirable data and processing overheads into the system.